Risk factors for pancreatic cancer mortality: Extended follow-up of the original Whitehall Study

Given the well-established links between diabetes and elevated rates of pancreatic cancer, there are reasons to anticipate that other markers of metabolic abnormality (increased body mass index, plasma cholesterol, and blood pressure) and their correlates (physical activity and socioeconomic status) may also confer increased risk. However, to date, the results of a series of population-based cohort studies are inconclusive. We examined these associations in the original Whitehall cohort study of 17,898 men. A maximum of 38 years of follow-up gave rise to 163 deaths due to carcinoma of the pancreas. Although Poisson regression analyses confirmed established risk factor-disease associations for increasingage, smoking, and type II diabetes, there was essentially no evidence that body mass index (rate ratio, 1.01; 95% confidence interval per 1 SD increase, 0.86-1.18), plasma cholesterol (0.91; 0.78-1.07), diastolic blood pressure (0.93; 0.78-1.09), systolic blood pressure (0.98; 0.83-1.15), physical activity (sedentary versus high: 1.37; 0.89-2.12), or socioeconomic status [clerical (low) versus professional/executive, 0.95; 0.59-1.51] offered any predictive value for pancreatic cancer mortality. These results were unchanged following control for a range of covariates.